Clinical Outcomes of ¹⁷⁷Lu-PSMA Radioligand Therapy in Earlier and Later Phases of Metastatic Castration-Resistant Prostate Cancer Grouped by Previous Taxane Chemotherapy

DISCUSSION

¹⁷⁷Lu-PRLT is a promising new treatment option for mCRPC, with several recent studies reporting high clinical efficacy and minimal associated toxicity even in patients with advanced disease heavily pretreated with multiple lines of systemic therapy (12–18,20). Our results support the growing evidence of good clinical efficacy of ¹⁷⁷Lu-PRLT in this patient group, with a median OS of 10.7 mo achieved in T-pretreated patients with advanced disease, most of whom had also received multiple additional treatments including enzalutamide or abiraterone. In addition, we demonstrated promising outcomes in T-naïve patients with less advanced disease achieving a median OS of 27.1 mo in our cohort. Although the T-naïve cohort only demonstrated a trend toward lower Gleason scores, it is possible that these patients also had less aggressive tumors from initial diagnosis given that T-pretreated patients had exhausted multiple additional lines of therapy in a similar time period before starting ¹⁷⁷Lu-PRLT. The overall results, however, suggest that ¹⁷⁷Lu-PRLT is a valid treatment option not only in heavily pretreated patients with mCRPC, but also in T-naïve patients with less advanced disease if deemed unsuitable for or who decline chemotherapy.

While newer hormonal agents are more recently being introduced as first-line therapies, docetaxel as first-line chemotherapy has been the mainstay of treatment in patients with mCRPC ever since demonstration of a survival benefit in 2 landmark randomized controlled trials published in 2004 (5,6). In the TAX 327 and SWOG 9916 trials, median OS results of 18.9 and 17.5 mo were seen in patients receiving docetaxel treatment regimens, resulting in a survival benefit of 2–3 months compared with other treatments (5,6). More recent studies have reported higher median OS when investigating newer taxane-based treatment regimes in chemotherapy-naïve mCRPC patients. Published in 2017, the FIRSTANA trial compared 2 different cabazitaxel regimes to docetaxel as first-line therapy in chemotherapy-naïve mCRPC, demonstrating a similar median OS of 24–25 mo in the 3 groups (28). Our retrospective series demonstrates a promising median OS of 27.1 mo in T-naïve patients treated with ¹⁷⁷Lu-PRLT. In addition, 38% of patients in our T-naïve cohort had been pretreated with abiraterone or enzalutamide compared with 2% in the FIRSTANA study. While definitive conclusions cannot be made between the efficacy of ¹⁷⁷Lu-PRLT compared with taxane-based chemotherapy by comparing our retrospective results to outcomes in different patient cohorts, our findings are encouraging and prospective clinical trials comparing the efficacy of ¹⁷⁷Lu-PRLT with initial taxane chemotherapy regimes in mCRPC appear warranted. Notably, high baseline ALP levels were an independent predictor of inferior OS in our T-naïve cohort treated with ¹⁷⁷Lu-PRLT, suggesting the burden of bone metastatic disease may be a particularly important factor in treatment decision making in earlier phases of the disease process.

Second-line therapeutic options following progression of mCRPC after docetaxel chemotherapy include cabazitaxel, newer hormonal agents and ²²³Ra. OS with these agents as second-line therapies in selected patient cohorts ranges from 14 to 18 mo (4,8,9,11). When progression occurs following these second-line therapies, treatment options are limited. ¹⁷⁷Lu-PRLT offers an additional therapeutic option in this phase of the disease, with our results demonstrating a median OS of 10.7 mo in T-pretreated patients, most of whom had also received abiraterone or enzalutamide. While these results are comparable to outcomes in the recently published COMET-1 trial where a median OS of 11.0 mo was reported in taxane- and abiraterone-pretreated patients treated with the kinase inhibitor cabozantinib, the vast majority (88%) of patients in COMET-1 had Eastern Cooperative Oncology Group performance status of 0 or 1 and there appeared to be fewer patients with Gleason scores of 8 or greater, suggesting that our T-pretreated cohort possibly had more advanced and aggressive disease (29). Therefore, our results lend support to recent evidence suggesting that ¹⁷⁷Lu-PRLT remains a promising therapeutic option in these treatment-refractory patients with advanced disease (30), particularly given that best supportive care is the typically recommended management in this late phase of the disease (3). In our study, independent predictors of inferior OS with ¹⁷⁷Lu-PRLT in this patient group were lower cumulative administered activity, lower baseline hemoglobin levels and poorer performance status highlighting the need to consider radiation dosimetry, laboratory results and overall patient condition in a personalized nuclear medicine treatment model in patients with advanced cancer. It is important to recognize that other predictors such as PSA doubling time, change in bone scan index, and lactate dehydrogenase may also be useful, although these were not all routinely available in our study population and were therefore unable to be assessed as independent predictors.

rPFS was superior in T-naïve patients compared with T-pretreated patients (8.8 vs. 6.0 mo), likely reflecting the more aggressive and advanced nature of the disease in the T-pretreated cohort. In T-naïve patients, the presence of visceral metastatic disease and higher baseline ALP levels were significant independent predictors of inferior rPFS in keeping with our clinical experience of patients with liver or lung metastatic disease or extensive bone metastatic disease being at higher risk for early disease progression. Although the rPFS results in this study are shorter than those previously reported by our institution, this is explainable by the use of molecular imaging criteria in the current study being more sensitive in detecting disease progression than the RECIST 1.1 used in our prior analysis (17). Our use of molecular imaging response criteria also makes comparison of rPFS results with other studies difficult given most have used RECIST-based criteria. Our PSA response rates appear comparable to the results of taxane-based chemotherapy trials (4–6,28). This also lends support to the high clinical efficacy of ¹⁷⁷Lu-PRLT, although relying on PSA as a primary measure of disease burden and treatment response may be imperfect given that dedifferentiated disease can be associated with minimal PSA elevation despite maintaining high PSMA expression (17).

Our results demonstrate minimal toxicity of ¹⁷⁷Lu-PRLT with grade 3 or 4 anemia, leukocytopenia, and thrombocytopenia seen in only 3%, 1%, and 2% of additional patients in the entire cohort, respectively. Notably, these minimal hematologic toxicities are overall lower than those seen with first- and second-line taxane-based chemotherapeutic agents (4–6,28). In particular, the rate of grade 3 or 4 leukocytopenia appears dramatically less with ¹⁷⁷Lu-PRLT potentially translating to a lower risk of treatment-related complications such as neutropenic septicemia with possible lethal outcomes (4–6,28). Furthermore, only 1 additional patient in the entire cohort achieved grade 3 renal toxicity as measured by eGFR. This patient had normal renal function before ¹⁷⁷Lu-PRLT, but suffered coexistent hypercalcemic nephropathy in the setting of extensive osteoblastic metastases, which was established as the cause for deterioration in renal function. The minimal hematologic and renal toxicity associated with ¹⁷⁷Lu-PRLT in this study is in keeping with prior results from our institution as well as other centers, providing further evidence for the safety of ¹⁷⁷Lu-PRLT even in patients with advanced disease (13,17,18).

Limitations of our study include its retrospective design and the heterogeneous nature of both patient cohorts including the prior treatments received. However, this analysis provides detailed outcomes of patients referred to our center in a real-world clinical setting and, therefore, we believe the findings are highly relevant to current clinical practice. The univariate/multivariable analysis is also likely to be limited by the small sample size in each cohort, which resulted in higher uncertainty in estimated HR for T-pretreated patients with bone metastases and T-naïve patients with high baseline ALP. In addition, there is possible bias introduced when using cumulative administered activity as a predictor of OS since patients who respond well to initial therapy are more likely to live longer and receive further ¹⁷⁷Lu-PRLT cycles. In our study, rPFS was based on European Organization for Research and Treatment of Cancer molecular imaging criteria, although these have not been validated for use with ⁶⁸Ga-PSMA PET/CT. Nevertheless, given that RECIST- and bone scan–based criteria have major limitations in assessing disease response in patients with extensive bone metastases, our results remain meaningful despite the absence of established clinical guidelines in this regard (31). Furthermore, PET-based molecular imaging forms the cornerstone of assessing suitability for radionuclide treatment and may be a preferred method of response assessment according to the principles of theranostics. Our PSA response assessment criteria of greater than 50% was based on prior guidelines (32) to allow comparison with earlier studies, although we note that recent evidence suggests using a lower PSA percentage decline may be a more optimal measure of ¹⁷⁷Lu-PRLT response (30).